…oyancy

Remove unused imports/variables/functions, fix mistakes, simplify
expressions, replace some magic numbers by expressions, replace
sys.stdout() by print(), remove commented code.

More cleanup and apply code review feedback: remove max_num_cells,
revert espressomd.System and box_l changes, use str.format().

Replace all occurrences of custom force_cap-based warmup code by
the steepest descent algorithm. Initialize the thermostat after
steepest descent. Don't print LJ interaction parameters again
after minimization.

Remove code writing out (sometimes incomplete) Tcl files. Write out
observables in tab-separated values files suitable for np.loadtxt()
and xmgrace plotting. Remove clutter from print() statements.

End links with a double underscore to make anonymous references and
avoid overshadowing explicit references (here the title reference).

… comm code.

        core/testsuite: engine_lb test fix from christophlohrmann

3277: Removed ENGINE shear torque calculation r=RudolfWeeber a=fweik

@RudolfWeeber @christophlohrmann is this what you had in mind?
If so the simplification is quite significant because all of the custom
communication code for ENGINE can be removed. Also the LB coupling
should now work in parallel.

Description of changes:
 - Removed ENGINE shear torque calculation
 - Moved swimmer parameters into ParticleProperties
 - Removed custom communiication code for ENGINE
 - Adaped docs


Co-authored-by: Florian Weik <[email protected]>

3282: utils: Re-add uniform test + docstring r=jngrad a=fweik

Description of changes:
 - The Utils::uniform test was moved to  the correct place and
   readded to the build system.
 - Docstring


Co-authored-by: Florian Weik <[email protected]>

Reference solutions for forces on shear planes and for stokes drag need dynamic viscosity, but LB uses kinetmatic viscosity.
This switches the tests to dynamic viscosity and removes a 1/rho from the unit conversion in the script interface.
Also, a force balance test is added, which does not depend on the density.

3267: Upgrade samples r=RudolfWeeber a=jngrad

Fixes #2089

Description of changes:
- remove unused variables/imports/functions and some of the commented out code
- remove Tcl-specific code, replace manual warmup by steepest descent
- use `argparse` to process user input and document the parameters, stop printing features
- improve docstrings and use them to generate an always up-to-date sample list in the docs


Co-authored-by: Jean-Noël Grad <[email protected]>
Co-authored-by: RudolfWeeber <[email protected]>

3287: LB Boundary force: fix unit conversion and tests r=fweik a=RudolfWeeber

Reference solutions for forces on shear planes and for stokes drag need dynamic viscosity, but LB uses kinetmatic viscosity.
This switches the tests to dynamic viscosity and removes a 1/rho from the unit conversion in the script interface.
Also, a force balance test is added, which does not depend on the density.

Fixes #

Description of changes:
 - 


PR Checklist
------------
 - [ ] Tests?
   - [ ] Interface
   - [ ] Core 
 - [ ] Docs?


Co-authored-by: Rudolf Weeber <[email protected]>

…oyancy

switch conditionals as suggested by JNGrad

Co-Authored-By: Jean-Noël Grad <[email protected]>

3288: core: Simplify particle removal r=RudolfWeeber a=fweik

Fixes #3286.

Description of changes:
 - Simplify particle removal by doing one linear pass over all particles
   instead using index


3290: fix writevtk() bug where types=all was ignored r=jngrad a=christophlohrmann

Description of changes:
 - fixed bug where the argument ``types='all'`` was ignored by ``system.part.writevtk()`` 


PR Checklist
------------
 - [ ] Tests?
   - [ ] Interface
   - [ ] Core 
 - [ ] Docs?


Co-authored-by: Florian Weik <[email protected]>
Co-authored-by: Christoph Lohrmann <[email protected]>
Co-authored-by: christophlohrmann <[email protected]>

2748: Additional LB test cases r=fweik a=RudolfWeeber

* Buyancy force on a sphere in a box of fluid under gravity
* Momentum conservation in a fluid with an external volume force applied and a particle with an oposite force applied

@KaiSzuttor @mkuron Could you please take a look at the
lb_buoyancy_force test? It only passes, when the fluid density is set to 1.
On the other hand, the lb_shear and lb_stokes_sphere tests check boundary forces which do not (and should not) depend on the density. They pass when run with different densities.
Also, the application of an external volume force is tested in lb.py in test_a_ext_force...
Furthermore, to the lb_momentum_conservation.py test checks the balance between forces applied to the lb and those applied to a particle.



Co-authored-by: Rudolf Weeber <[email protected]>
Co-authored-by: Florian Weik <[email protected]>

… removed special overload for boost::optional

POSIX does not specify test for > 4 arguments:
https://pubs.opengroup.org/onlinepubs/9699919799/utilities/test.html

3291: Factor out and test memcpy serialization r=RudolfWeeber a=fweik

Description of changes:
 - This splits off the memcpy serialization used in the ghost communication and adds unit test
   for it.
 - Added customization point to opt-in to memcpy serialization for types that are not [TriviallyCopyable](https://en.cppreference.com/w/cpp/named_req/TriviallyCopyable), but serializes
 to a bounded size, like `boost::optional`. This is so we can use `optional` and `variant` in `ParticleProperties`, which is planned.


3292: Pypresso: Remove "test -a" and quote $@ r=jngrad a=hirschsn

Description of changes:
 - Remove `test -a` since calls to test with more than 4 arguments are unspecified: https://pubs.opengroup.org/onlinepubs/9699919799/utilities/test.html
 - Quote `$@` to avoid splitting $@. Currently `./pypresso --gdb "test me.py"` will fail because tries to find the file `test`. Quoting avoids this.


PR Checklist
------------
 - [ ] Tests?
   - [ ] Interface
   - [ ] Core 
 - [ ] Docs?


Co-authored-by: Florian Weik <[email protected]>
Co-authored-by: Steffen Hirschmann <[email protected]>